Helpful answers

Frequently asked questions

How the products work, what they will and won't do without you, how your data is handled, and where early access and pricing stand.

General

What is AI4Everyone?

A set of products that each act as an AI team for one job: Video makes YouTube videos and Shorts, Social runs your brand's Instagram, and Product Reviewer tests your website the way your customers use it. You set the goal and the limits, the AI does the work and checks it, and you make the final call.

Who is it for?

Creators running YouTube channels, small brands and the agencies that manage their social accounts, and teams who want to know how their website really performs for customers. If you have more work than hands but still want the final say on what goes out, it's built for you.

Will it ever post or publish without me?

No. The AI proposes, and a person approves. Every YouTube upload, every Instagram post and reply, and every step that would submit, buy or delete something on a website waits for your approval. You can schedule what you've approved, but nothing reaches the outside world unseen.

Which AI models does it use?

Whichever does the job best at the price you've chosen. You pick outcomes like quality, speed or cost, not models. Every model provider sits behind an interface we own, so models can be swapped as better ones arrive without changing anything for you.

Video

What does Video make?

It turns your sources into ideas, ideas into scripts, and scripts into finished YouTube videos and Shorts, organised by channel and series. One script can become several independent cuts, such as a 60-second Short, a longer explainer and a Hindi version.

Where do the ideas come from?

From the places you point it at: subreddits, articles, PDFs, your own notes and footage. Every idea arrives with its hook, the sources it came from, and a check against what your channel has already covered.

Is this a way to flood YouTube with AI videos?

No, and it's designed not to be. Each channel has a conservative daily publishing limit (two videos a day by default) that you can't raise from the app, a new video that is too similar to your recent ones is blocked before it's queued, and every upload needs your own approval. YouTube penalises mass-produced, repetitive content, and the penalty would land on your channel, so we treat volume as a risk rather than a feature.

Will my videos be labelled as AI-generated?

Yes, by default. Video turns on YouTube's altered or synthetic content disclosure for every upload. Disclosing doesn't limit a video's reach or its eligibility to earn money, and leaving it off where it's required can get a video removed. You can switch it off only where the content genuinely isn't realistic, and that choice is recorded.

Can I fix one scene without remaking the whole video?

Yes. You can redo a single shot, caption or hook. The rest of the video stays as it was, and only the part you changed is made again.

Can it make an AI presenter who gives health, legal or financial advice?

No. YouTube won't let channels monetise AI personas on sensitive topics such as health, medicine, law, finance and politics, so Video refuses to create one and tells you it's YouTube's rule. Factual videos on those subjects without a synthetic presenter are fine.

Social

Which social networks does Social support?

Instagram first: publishing and scheduling posts and reels, comments, DMs and analytics. Facebook, WhatsApp, YouTube, LinkedIn, Google Business and X are planned next.

What kind of Instagram account do I need?

A Business or Creator account, because Instagram only allows those to be managed through its API. If you have a personal account, Social walks you through switching it before connecting.

I don't have an Instagram account yet. Can you create one?

We guide you through setting one up, but you create it yourself on Instagram's own sign-up pages. We never create accounts by automation, and we never see or handle your Instagram password. Once your new account is a Business or Creator account, you connect it like any other.

Does it reply to comments and DMs on its own?

No. It gathers comments and DMs into one inbox and drafts replies in your brand's voice. You edit or approve each reply before it's sent.

What happens if my Instagram connection breaks?

Connections are checked regularly. If one degrades or needs reconnecting, you're told and prompted to fix it, rather than finding out later that posts quietly stopped going out.

Product Reviewer

What does Product Reviewer do?

Give it your web address. It works out what you sell and who your customers are, then walks their journeys in a real browser on desktop and mobile. You get findings backed by screenshots and steps, suggested fixes, and a way to check that each fix worked.

Do I have to write a checklist or test cases?

No. There's no predefined checklist and no score out of a hundred. Your website decides what's relevant: Reviewer infers the customers, goals and journeys from the site itself. You can review and adjust what it found, or ask it a specific question such as "why might people abandon checkout?"

Can I point Product Reviewer at any website?

Only your own site, or one you have written permission to test. That's a condition of use in our terms, not a suggestion.

Will it place real orders or send real messages on my site?

It stops before anything irreversible, such as paying, placing an order, sending a support request or deleting data. It uses test or sandbox paths where your site has them, asks you before going further, and labels simulated completions clearly so they're never mistaken for real ones.

Can it test pages behind a login?

Yes. You can store test credentials for different kinds of customer. They're encrypted and never shown again after you save them. When a site needs a one-time code, a passkey or a CAPTCHA, Reviewer hands control to you to complete it and then carries on.

How do I know a fix actually worked?

Mark a finding as ready and Reviewer re-runs the journey that found it and tells you whether it's fixed. Scheduled re-checks can then keep watching for regressions.

Chat & agents

Can I use it from Claude, ChatGPT or my own agent?

That's a core design goal. Every product is designed MCP-first: the plan is one MCP server per product plus one for the whole platform, so every capability is reachable from chat and from AI agents, not only from the web app. Agent access is planned to open alongside each product.

Can an AI agent change things without my approval?

No. Anything an agent asks for that would change, send or spend something comes back as a proposal with a plain-language summary. Nothing runs until a person approves it, and the action is recorded.

What stops a web page from giving the AI instructions?

Text the AI reads from outside, such as a chat message or a page Product Reviewer crawled, is treated as data, never as instructions. That boundary is one of the reasons every write is a proposal a person approves.

Privacy & security

What does signing in with Google give you access to?

Only who you are: your name, email address and a stable account identifier. It grants no access to YouTube, Drive, Gmail or any other Google service. Connecting a YouTube channel or an Instagram account is a separate step you take only when you need it, and you can decline it and still use the rest of the product.

Do you sell my data or use it to train AI models?

No to both. We share data only with the infrastructure and services needed to run the platform, and with the services you explicitly connect, to the extent needed to do what you asked. See our privacy policy.

How is my data kept separate from other customers'?

By the database itself, not just by application code. Every record carries the workspace that owns it, and the database refuses to return rows from another workspace even if the application asks. Each product also keeps its own data, separate from the others.

How are my connected accounts and passwords protected?

Credentials for connected services are encrypted before they're stored and decrypted only to carry out an action you asked for. Session tokens are stored only as a hash, so a copy of our database wouldn't let anyone sign in as you.

Where is my data stored, and can I delete it?

Data is processed and stored on infrastructure in the United States. You can disconnect a service at any time, and you can ask us to delete your account and its data by writing to support@ai4everyone.dev.

Who owns what the products make?

You do. You keep ownership of everything you upload and everything the products generate for you. We only get the permission needed to run the service for you, and it ends when you delete the content or your account.

Access & pricing

How much does it cost?

Pricing hasn't been published yet. It will be published before anything is charged, and nothing is charged before then. Inside the products you'll always see an estimate before any work that costs money starts.

When can I start using it?

Early access opens one product at a time. Tell us which product you want first and we'll let you in as it opens. We haven't announced dates. Request early access.

Will I know what something costs before it runs?

Yes. Anything that costs money shows an estimate first, stays inside the cap you set, and shows what it actually cost afterwards. Where the exact cost can't be known in advance, as with a website investigation, you see a range and a hard limit instead of false precision. If a job reaches your cap it pauses, keeps the work already done, and lets you decide whether to raise the limit.

Company

Who is behind AI4Everyone?

AI4EVERYONE LLP, an Indian AI startup based in Visakhapatnam, Andhra Pradesh. More about us.

Is AI4Everyone a registered company?

Yes. AI4EVERYONE LLP was incorporated on 25 February 2026 (LLPIN ACV-7809), is recognised as a startup by DPIIT under Startup India (certificate DIPP246594, valid to 24 February 2036), and is registered for GST (GSTIN 37ACNFA4617L1Z0).

Still have a question?

Write to support@ai4everyone.dev and we'll get back to you.